(1) Field of the Invention
The present invention relates to cDNA encoding polypeptides which when attached to the C-Terminal region of a protein enables sorting of proteins to the vacuoles in plants. In particular, the present invention relates to polypeptides which enables the sorting of lectins, which are insecticidal polypeptides, to the vacuole of a plant.
(2) Prior Art
In eukaryotes, proteins of the endoplasmic reticulum (ER), Golgi, lysosomes, vacuoles, plasma membrane, and cell wall are derived from a subset of proteins that enter the secretory pathway. Proteins are targeted to the secretory pathway by an N-terminal hydrophobic signal sequence which mediates a transmembrane translocation from the cytosol to the lumen of the endoplasmic reticulum. Following proteolytic cleavage of the signal sequence, some secretory proteins undergo further post-translational processing in the ER and Golgi network (Blobel, G., and Dobberstein, D., J. Cell Biol. 67, 835-851 (1975)). Proteins traversing the secretory pathway are believed to be sorted to their respective compartments by selective retention or targeting information contained in their molecular structures (Rothman, J. E., Cell 50, 521-522 (1987)). Proteins lacking specific sorting determinants follow a default pathway and are consequently secreted toward the cell surface (Rothman, J. E., Cell 50, 521-522 (1987); Wieland, F. T., et al., Cell 50, 289-300 (1987); Dorel, C., et al., J. Cell Biol. 108, 327-337 (1989); and Denecke, J., et al., Plant Cell 2, 51-59 (1990)).
A secondary sorting signal that mediates a targeting process involves either a post-translational modification of the protein or depends upon primary, secondary, or tertiary structural elements within the polypeptide (Verner, K., et al., Protein translocation across membranes. 241, 1307-1313 (1988)). The most well characterized sorting process is the mannose-6-phosphate dependent sorting of mammalian lysosomal enzymes (Kornfeld, S., et al., Ann. Rev. Cell Biol. 5, 483-525 (1989)). The active sorting of these enzymes to the lysosome is dependent on the modification of specific glycans with mannose-6-phosphate and the binding of the modified glycan to mannose-6-phosphate receptors in the trans-Golgi network (reviewed in Kornfeld, S., and Mellman, I., Ann. Rev. Cell Biol., 5, 483-525 (1989)). However, there is evidence for the existence of a mannose-6-phosphate independent system for the sorting of some mammalian lysosomal enzymes (Gabel, C. A., et al., Proc. Natl. Acad. Sci. U.S.A. 80, 775-779 (1983)).
In yeast and plants, N-linked glycans are not necessary for the correct transport and sorting of secretory proteins to vacuoles (Stevens, T. H., et al., Cell 30, 439-448 (1982); Voelker, T. A., et al., Plant Cell 1, 95-104 (1989); Wilkins, T. A., et al., Plant Cell 2, 301-313 (1990); and Sonnewald, U., et al., Plant Cell 2, 345-355 (1990)). Therefore, it appears that targeting of proteins to vacuoles in yeast and plants is independent of post-translational modifications to oligosaccharide side-chains and may be dependent upon elements within the polypeptide. Such peptide sorting determinant is identified for the yeast vacuolar carboxypeptidase Y (CPY). It has been demonstrated that the amino-terminal propeptide of CPY fused with the secreted enzyme invertase, contains the sorting signal of CPY (Johnson, L. M., et al., Cell 28, 875-885 (1987); and Valls, L. A., et al., Cell 48, 887-897 (1987)). A detailed mutational analysis of the amino-terminal propeptide determined that the tetrapeptide QRPL functions as a vacuolar sorting signal. Interestingly, the context in which the QRPL sequence is presented affects the efficiency of targeting, inferring the involvement of secondary structural elements in the sorting mechanism of CPY (Valls, L. A., et al., J. Cell Biol. 111, 361-368 (1990)). A sorting determinant was identified in the amino-terminal propeptide of another yeast vacuolar enzyme, proteinase A (Klionsky, D. J., et al., Mol. Cell Biol. 3, 2105-2116 (1988)), which is sufficient to redirect the normally secreted enzyme invertase to the yeast vacuole. However, currently no consensus sequence or common structural determinant has been demonstrated for targeting of yeast vacuolar proteins, suggesting that a diverse array of factors are involved in the sorting process.
It has been also shown that the plant vacuolar protein phytohemagglutinin-L (PHA), a lectin of Phaseolus vulgaris, is correctly processed and sorted to the yeast vacuole (Tague, B. W., et al., J. Cell Biochem. Suppl. 13D, 230 (1989)). Deletion analysis of PHA localized the vacuolar sorting domain within the amino-terminal portion of mature PHA (Tague, B. W., et al., Plant Cell 2, 533-546 (1990)). This domain contains a yeast-like targeting tetrapeptide sequence (LQRD), that is sufficient to target PHA-invertase hybrid proteins to the yeast vacuole (Tague, B. W., et al., Plant Cell 2, 533-546 (1990)). It should be noted however, that the same PHA-invertase fusion proteins were not successfully targeted to vacuoles in Arabidopsis thaliana protoplasts (Chrispeels, M. J., Ann. Rev. Plant Physiol. Plant Molec. Biol. 42, (1991)). Therefore, this sorting determinant contains enough information for vacuolar sorting in yeast, but appears to lack the necessary information for efficient targeting in plants, suggesting that vacuolar sorting signals in yeast and plants are dissimilar.
The ability to sort proteins to the vacuoles of plants is very useful. The result is that the fruits or leaves contain more protein which for edible plants is important to the food value of the plant or which contain antimicrobial agents such as lectins which are active against fungi and insects which attack the plant.
Several publications describe the sequencing of the lectins and the cDNA encoding the lectins for wheat, barley and rice. These publications are Raikhel, et al., Proc. Natl. Acad. Sci. 84, 6745-6749 (1987); Smith et al., Plant Molecular Biology 13, 601-603 (1989); Plant Physiology 91, 124-129 (1989); and Wilkins et al 1, 541-549 (1989). The structure of the lectins expressed by the genes was described with the peptide residues described herein; however, no utility for the residues was disclosed. The residues were not isolated from the lectin protein encoded by the cDNA.